RSS Feeds
Public Health Surveillance for Hereditary Hemochromatosis
- Scott F. Wetterhall, MD, MPH;
- Mary E. Cogswell, DrPH, RN; and
- Kris V. Kowdley, MD
- From the Centers for Disease Control and Prevention, Atlanta, Georgia; and University of Washington, Seattle, Washington. Requests for Reprints: Scott F. Wetterhall, MD, MPH, Mailstop D-24, Centers for Disease Control and Prevention, 1600 Clifton Road NE, Atlanta, GA 30333. Current Author Addresses: Dr. Wetterhall: Mailstop D-24, Centers for Disease Control and Prevention, 1600 Clifton Road NE, Atlanta, GA 30333; e-mail, sfw1{at}cdc.gov. Dr. Cogswell: Centers for Disease Control and Prevention, Mailstop K-25, 1600 Clifton Road NE, Atlanta, GA 30333. Dr. Kowdley: Department of Medicine, University of Washington, Box 356424, Seattle, WA 98196. Note: This article is one of a series of articles comprising an Annals of Internal Medicine supplement entitled “Iron Overload, Public Health, and Genetics.” To view a complete list of the articles included in this supplement, please view its Table of Contents.
Abstract
The recent realization that hemochromatosis is a common condition has created opportunities to develop unified public health surveillance for this disorder and its complications and to design programs to prevent unnecessary illness and death resulting from this disorder.Public health surveillance for hemochromatosis can be used to measure the magnitude of the problem (for example, to establish the number of persons with evidence of early iron overload); identify research needs; reveal the natural history of the disease; detect changes in health care practices, such as use of screening tests; and evaluate interventions, such as phlebotomy. Existing surveillance has been limited to periodic measurement of morbidity and mortality done by using hospital discharge records, health examination surveys, vital statistics, and data from small research registries. The improvement of surveillance will entail the ongoing collection of information from population-based surveys, such as the Behavioral Risk Factor Surveillance System; the collection of data on provider practices (for example, through the National Ambulatory Medical Care Survey); and the establishment of population-based registries. Creating population-based registries requires consensus on case definitions; strategies to encourage case ascertainment and reporting; policies and procedures for protecting privacy and ensuring confidentiality; and partnerships among providers, researchers, and public health officials. Longitudinal data from population-based registries will provide insight into determinants of disease expression, such as pattern or degree of iron overload. This information is critical for developing evidence-based recommendations for population screening, monitoring changes in medical practices, and assessing the effect of preventive measures.
The discovery that HLA-linked hereditary hemochromatosis is common (it is estimated to affect 1 in every 200 to 400 persons in the United States [1, 2]) and the recent discovery of the gene for this disorder [3] have created new challenges. Available data suggest that life expectancy is normal if hereditary hemochromatosis is recognized before cirrhosis appears but that it is decreased if cirrhosis has already developed, even if patients receive iron depletion therapy [4]. Thus, some experts have advocated the use of screening programs so that treatment can be started before clinical signs develop [5, 6]. Additional information, however, is needed to justify screening and to develop cost-effective screening programs.
Public health surveillance for any condition involves the ongoing and systematic collection, analysis, and interpretation of health data and the timely dissemination of these data to those responsible for preventing and controlling the condition [7-9]. Surveillance data are used to formulate public health policies, identify high-risk populations, target interventions, and evaluate progress in disease prevention [10]. No unified surveillance system exists for hereditary hemochromatosis. We describe the objectives of such a surveillance system, examine existing screening practices, propose enhancements to these practices, and discuss special issues in surveillance for hereditary hemochromatosis.
Objectives of Public Health Surveillance
Because its primary emphasis is on research and on prevention of disease through screening and treatment, public health surveillance for hereditary hemochromatosis has the following objectives: to measure the magnitude of the problem (for example, to establish the number of homozygotes or the number of persons undergoing phlebotomy treatment), identify research needs, improve understanding of the natural history of the disease, identify changes in screening practices by health care providers (for example, changes in compliance with recommendations for screening or appropriate use of screening tests), and evaluate the effect of prevention and control measures on the severity of disease.
Existing Practices
The limited efforts in public health surveillance for hereditary hemochromatosis to date have been directed at measuring the prevalence of iron overload disorders and identifying research needs. Surveillance has also provided some understanding of the natural history of hereditary hemochromatosis and the usefulness of diagnostic tests in identifying persons with this condition. This information has come from morbidity surveys, mortality data, and disease registries (Table 1).
Morbidity Surveys
The National Hospital Discharge Survey (NHDS) is a continuous, voluntary survey of noninstitutional, short-stay hospitals that provides national estimates of the number of hospitalizations that are associated with particular medical conditions (which are coded by using the most recent version of International Classification of Diseases [ICD]) [11, 12]. Advantages of the NHDS include its long period of operation (which allows examination of temporal trends) and its inclusion of information on comorbid conditions (data for each hospitalization are collected on as many as seven conditions and three procedures). The NHDS, however, cannot generate state-specific estimates; it enumerates hospital discharges rather than individual patients (that is, one person can contribute more than one hospitalization); and, for rare or underdiagnosed conditions (such as hereditary hemochromatosis), it results in diagnoses of questionable accuracy. In addition, the ICD-9-CM (ICD, 9th Revision, Clinical Modification) code used for hereditary hemochromatosis (275.0, disorders of iron metabolism) may also be used for secondary forms of hemochromatosis. Unless diagnoses with code 275.0 that also have the ICD-9-CM codes associated with secondary forms of hemochromatosis (such as 282, hereditary hemolytic anemias) are excluded during analysis, the use of code 275.0 will produce inaccurate estimates of the number of hospitalizations for hereditary hemochromatosis [17].
Medicaid claims data were used to estimate the hospitalization rate for hemochromatosis in 1984 in four states [11]. The advantages of claims data include their representativeness (for example, Medicare data include almost all persons 65 years of age or older), the opportunity they provide to analyze longitudinal information on individual patients (because each patient has a unique personal identifier), and the fact that they allow the calculation of rates for certain population subgroups (if denominator files of beneficiaries are available) [24-27]. Despite these advantages, the size and complexity of claims data systems, their differences across states, confidentiality issues, the difficulties of using an outpatient procedure classification system, and problems in distinguishing incidence from prevalence need to be addressed before claims data can be systematically used to conduct surveillance for hereditary hemochromatosis [24, 26, 27].
The National Health and Nutrition Examination Survey (NHANES) is a nationally representative survey that periodically collects information obtained from interviews, physical examinations, and laboratory tests [28]. The NHANES III could be used to estimate the U.S. prevalence of the gene for hereditary hemochromatosis; to examine the association of genetic risk factors with clinical and biochemical findings; and to gather data on risk factors, such as iron supplementation [13-15]. The NHANES has a complex sample survey design that allows only national, not state-specific, estimates. The use of different methods for measuring iron status in successive cross-sectional surveys (NHANES II and NHANES III) limits temporal comparisons [16]. In NHANES, a single transferrin saturation test is used without follow-up confirmation; thus, prevalence estimates for hemochromatosis may be overestimated, particularly in persons who provide nonfasting blood samples. Because of their size, cost, and complexity, these national surveys are infrequent.
Mortality Data
Data on vital statistics benefit from an established classification and coding system (disorders of iron metabolism have had their own ICD code [ICD-9 275.0] since 1968) and complete population coverage. Use of a model death certificate provides some standardization of data collection procedures across states, but limited information on potential risk factors, particularly risk factors associated with hemochromatosis (such as family history and use of iron supplements), is collected. The accuracy of cause-of-death data varies considerably, particularly for uncommon conditions such as hemochromatosis, which may go undetected or may be diagnosed on the basis of nonstandardized clinical criteria [18].
Disease Registries
Small-scale, hospital-based registries and case series have provided useful clinical and research information on hemochromatosis (such as the risk for liver cancer in patients with hemochromatosis and the lack of an association between hemochromatosis and coronary artery disease), but they cannot be used to calculate population-based rates [19-23]. Too often, referral patterns influence the size and composition of registries, making the calculation of valid population-based rates-and risks for disease-impossible. In addition, because registries actively collect a substantial amount of information from multiple sources over time, they tend to incur higher costs than other forms of surveillance [29].
Enhancing Public Health Surveillance
Information systems are needed to answer such questions as, How many persons are at risk for hemochromatosis? What is the natural history of the disease, particularly when the disease is detected early through screening of asymptomatic persons, and how can this history be modified? What is the distribution of risk factors for disease expression in the U.S. population? If the decision is made to implement population-based screening, we will need to ask further questions. Are physicians and the public aware of the screening guidelines? Are the appropriate persons being screened? What treatment and follow-up options are available for persons screened and found to be positive for hereditary hemochromatosis? What are the ethical, legal, and social implications of a positive screening test result? Data could be gathered in population-based surveys, surveys of physician practices, and population-based registries (Table 2).
Population-Based Surveys
Population-based surveys are a convenient way to monitor public awareness of hereditary hemochromatosis, the need for screening, and the distribution of factors that may influence risk for disease expression. The Behavioral Risk Factor Surveillance System (BRFSS) can be used to assess knowledge and attitudes about hemochromatosis in the U.S. adult population, measure behavioral factors (for example, the prevalence of iron supplement use) that can increase risk for iron overload, monitor the proportion of persons screened for hereditary hemochromatosis, and tailor health education messages to particular audiences [30]. However, the BRFSS collects self-reported data from telephone interviews, and it can be difficult to judge the accuracy of these data. Moreover, because the advantages of the BRFSS are numerous, there is competition to add questions to the survey.
Provider Surveys
To circumvent reliance on patient-reported data, questions can be added to existing surveys (such as the National Ambulatory Medical Care Survey [NAMCS], an ongoing national probability sample of medical encounters in ambulatory settings) to characterize the screening practices of physicians [31]. A question on iron overload screening could be used to estimate the annual number of iron screening tests performed and to target educational messages to physicians about the importance of screening for hereditary hemochromatosis.
Population-Based Registries
In operating population-based registries, staff members actively seek all new cases of a disease within a geographically defined area (for example, by contacting hospitals and health care providers) and obtain information on individual cases from many sources [32, 33]. Sources for new cases of hemochromatosis include hospitals, blood banks, laboratories, physicians, and registries of liver transplant recipients and patients with cancer. Data from a population-based hemochromatosis registry could be used to determine the incidence of the disease, conduct research, examine trends over time, and evaluate the effectiveness of screening and treatment programs.
Special Issues
Case Definitions
Case definitions for hemochromatosis, which currently lack consistency, are a uniform set of criteria (based on clinical, laboratory, and epidemiologic information) for classifying the presence of a disease or condition [34, 35]. The College of American Pathologists [36] proposes that the working definition of hereditary hemochromatosis be a transferrin saturation greater than 60% on two occasions with no other known cause for the elevation. Other researchers who advocate widespread screening for hereditary hemochromatosis suggest that a lower cut-off transferrin saturation (>50%) be used for women [4] or that unsaturated iron-binding capacity testing is adequate for screening [37]. A further complication is that such case definitions are based on what are generally accepted as screening criteria; additional tests, such as liver biopsy, quantitative phlebotomy, or genetic testing, are needed to confirm the diagnosis. Such confirmatory tests are frequently missing from case definitions.
Case definitions simplify and standardize reporting practices, thereby improving case ascertainment and data quality. Consistent case definitions permit accurate assessment of secular trends and enhance the study of risk factors for disease progression and expression. Nevertheless, the case definitions used in surveillance are usually not intended as the only criteria for clinical diagnosis [34]. Case definitions for an evolving or previously unrecognized condition often have many criteria. For the public health surveillance of hemochromatosis, we propose the following working definitions. A suspected case is defined by one elevated transferrin saturation measure; a probable case is defined by two elevated transferrin saturation measures; and a confirmed case is defined by at least 4 g of iron removed by quantitative phlebotomy or at least one of the following: grade 3 to 4 iron stores on Perls stain on liver biopsy, a hepatic iron concentration greater than 80 µmol per g dry weight on liver biopsy, or a hepatic iron index value of 1.9 or more on liver biopsy [38].
New developments, particularly laboratory advances, will force the revision of case definitions. With the recent discovery of two mutations in the putative gene for hereditary hemochromatosis [3] and the marketing of the gene test for commercial use, diagnoses now may be made on the basis of genetic tests. Although DNA-based testing may simplify case diagnosis, phenotypic screening will remain important. This is because many persons in whom hemochromatosis is diagnosed may not have a mutation and because it is unknown whether all persons who are genetically homozygous for the mutations will develop the disease [39].
Reporting Practices
Physicians have not routinely screened for hereditary hemochromatosis and have waited for the appearance of clinical signs and symptoms before initiating diagnosis and treatment [36]. Furthermore, hereditary hemochromatosis is widely believed to be rare. Thus, in most persons, the disorder is undetected until iron overload and clinical signs and symptoms develop [23]. As physicians become more aware of the importance of screening for hemochromatosis, the application of screening tests to the prevalent pool of undiagnosed cases may create the impression that the frequency of this disorder is increasing. Media attention [40], heightened awareness of a new medical condition [41], or changes in a case definition [42] may also prompt increased reporting of a condition to a surveillance system. Once reporting practices stabilize, however, surveillance for hereditary hemochromatosis will permit monitoring of the use of screening tests and early case detection in persons with the disorder.
Underreporting can result in unreliable comparisons across geographic or population subgroups, the introduction of bias (because the reported cases may not be representative of the broader spectrum of cases), and an underestimate of disease incidence [43]. Underreporting has two general causes: failure to diagnose and failure to report [44]. Failure to diagnose can result from interference with appropriate diagnosis-for example, lack of contact of a patient with the medical system, a physician's unfamiliarity with the laboratory methods available for detecting hereditary hemochromatosis, or a physician's lack of awareness of the early and nonspecific manifestations of disease [23]. Failure to report a diagnosed case to a disease registry or other public health authority is usually caused by lack of familiarity with the reporting process, concerns about confidentiality, or lack of time [45, 46].
Surveillance systems also need to devise strategies to encourage compliance with reporting. Improving public awareness, educating physicians, simplifying the surveillance system, and providing feedback will enhance participation. Alternatively, active surveillance, done by searching multiple data sources (such as laboratories and hospital discharge records), can be used to identify possible cases [47, 48].
Privacy and Confidentiality
For some types of surveillance activities (such as the BRFSS and the NAMCS), data can be collected anonymously and personal identifiers can be eliminated. For other types of surveillance (including NHANES and population-based registries), personal identifiers are retained for follow-up and the protection of privacy and confidentiality poses greater challenges. Concerns about privacy are heightened by uncertainties about testing for genetic conditions and concerns about genetic discrimination, stigmatization, and cost to the patient (for example, from loss of employment or denial of health insurance) [49-52]. Genetic testing will require informed consent, genetic counseling, and follow-up for persons whose tests results are positive [53]. Several professional associations have endorsed the use of informed consent for genetic testing in clinical and research settings and for use of stored tissue samples [54-56]. Other groups have encouraged the use of certificates of confidentiality, which are assurances issued under the Public Health Service Act that protect against compulsory disclosure (for example, disclosure under subpoena) [53, 54]. Privacy recommendations associated with the Health Insurance Portability and Accountability Act (Public Law 104-191) and several recent congressional initiatives that prohibit discrimination in health insurance and employment will, if enacted, give patients additional protection [57].
Involvement on the State and Local Level
Crucial to the success of the prevention and management of iron overload are the partnerships of the Centers for Disease Control and Prevention with other federal agencies; state and local health departments; professional organizations; academic institutions; private health care providers; the managed care, laboratory, and blood banking industries; local community organizations; and patient advocacy groups. Dedicated resources for surveillance on iron overload are limited, however, and programs for this surveillance compete with programs for surveillance of other conditions [58]. In addition, no federal resources are currently allocated to state and local health departments to support chronic disease program efforts related to hereditary hemochromatosis. Partnerships will reduce duplication of effort, but more resources are needed to establish an active, population-based surveillance system.
Recommendations
To enhance public health surveillance for hemochromatosis, we recommend the following steps:
1. Develop and agree on surveillance case definitions for hemochromatosis.
2. Improve current surveillance systems by improving awareness of hereditary hemochromatosis by health care providers (if providers recognize how common the disorder is, the frequency with which it is diagnosed will increase, resulting in more accurate surveillance data); including the appropriate tests in surveys with examination components, such as NHANES; developing, testing, and implementing questions about screening for hereditary hemochromatosis in population-based surveys, such as BRFSS; and including questions on screening for hereditary hemochromatosis in the NAMCS.
3. Balance the need to protect privacy with the social benefit of using information to further research and public health.
4. Create population-based registries for hemochromatosis.
5. Forge partnerships for surveillance among federal, state, and local health officials; members of academia; and medical practitioners.
Conclusions
Public health surveillance is a critical part of the strategy to prevent hereditary hemochromatosis. Surveillance for this condition complements efforts in public education, provider awareness, and applied research-the fundamental strategies for decreasing the burden caused by the largely preventable manifestations of this genetic condition. The challenge is to develop a public health surveillance system for monitoring this disease and its complications in an era of diminished resources.
The development of a national public health surveillance system for hereditary hemochromatosis will provide useful information to those wishing to establish monitoring systems for other genetic diseases for which knowledge of the determinants of phenotypic expression-and the need for clinical therapy-is still evolving. Agreement on case definitions, identification of areas for additional biomedical and population-based research, and acceptance of the opportunity costs associated with population-based monitoring are issues that must be faced during the set-up of screening and surveillance systems for any genetic disease. The lessons learned with hereditary hemochromatosis will help us address the pressing clinical, ethical, legal, and social issues related to screening for and monitoring other genetic disorders.
By engaging in public health surveillance for hereditary hemochromatosis, medical providers, researchers, and public health officials can better determine the burden of illness that results from this disorder. Longitudinal data from population-based registries will provide insight into the determinants of disease expression (manifested as iron overload), the course of disease during treatment, and factors influencing patient compliance with treatment. Such information is critical for developing evidence-based recommendations for population screening, monitoring changes in medical practices, and assessing the effect of preventive measures. Providing the foundation for the systematic collection and use of this information is challenging for hereditary hemochromatosis and other genetic diseases. However, the opportunity to prevent disease and improve quality of life for the estimated 1 million persons in the United States at risk for unnecessary illness or death from hereditary hemochromatosis is an opportunity that Americans cannot afford to ignore.
- Copyright ©2004 by the American College of Physicians
